1. Definition of Interest Rate

The interest rate $r$ can be interpreted as:

  1. A required rate of return: The investor’s required compensation.
  2. A discount rate: The rate used to determine the present value of future cash flows.
  3. An opportunity cost: The return foregone by investing in a particular asset instead of the next best alternative.

These reflect the relationship between cashflows.

2. The Interest Rate and Time Value of Money

$$\text{Interest rate} = \text{Risk-free rate} + \text{Premiums (bearing distinct types of risk)}$$

Also called the yield.

Example: $9500 today equals $10000 in one year → this is the discount rate.

$$\frac{10{,}000 - 9{,}500}{9{,}500} = \frac{500}{9{,}500} \approx 5.26% \quad \text{(rate of return)}$$

2.1 Determinants of Interest Rates

$$r = r_{RF} + \text{Inflation premium} + \text{Default risk premium} + \text{Liquidity premium} + \text{Maturity premium}$$

PremiumCompensates investors for…
Real risk-free rate $r_{RF}$Time preference only — assumes no inflation
Inflation premiumExpected (average) inflation rate
Default risk premiumPossibility that the borrower fails to make a promised payment at the contracted time and amount
Liquidity premiumRisk of receiving less than fair value if the investment needs to be converted to cash quickly
Maturity premiumLong-term bonds expose investors to more interest rate risk; investors are locked to a fixed rate even when better rates appear in the market

2.2 Nominal vs. Real Risk-Free Rate

The nominal risk-free interest rate reflects both the real risk-free rate and the inflation premium.

Fisher equation (exact):

$$1 + i = (1 + r_{RF})(1 + \pi)$$

$r_{RF}$: Nominal risk-free rate

$\pi$: Inflation rate

Approximation (when $r_{RF}$ and $\pi$ are small):

$$i \approx r_{RF} + \pi$$

since $(1 + r_{RF})(1 + \pi) = 1 + r_{RF} + \pi + r_{RF}\pi \approx 1 + r_{RF} + \pi$.

as $r_{RF}\pi$ is very small and can be ignored.

3. Rates of Return

Calculate and interpret different approaches to return measurement over time and describe their appropriate uses.

Financial assets generate returns through:

graph TD
    A[Financial Asset Return] --> B[Periodic Income]
    A --> C[Price Movement]
    B --> D[Cash dividends]
    B --> E[Interest payments]
    B --> F[Pension plans / retirement annuities]
    C --> G[Capital gain / loss]
    C --> H[Non-dividend-paying stocks]

3.1 Holding Period Return

$$R = \frac{P_1 - P_0 + I_1}{P_0}$$

where $P$ is the price and $I_1$ is the income; we buy at time 0 and sell at time 1.

For periods longer than 1 year, with annual returns $R_1, R_2, R_3$:

$$R = \left[(1+R_1)(1+R_2)(1+R_3)\right] - 1$$

3.2 Arithmetic Mean Return

For holding-period returns (daily, monthly, annual), the arithmetic mean (mean return) is:

$$\bar{R}i = \frac{R{i,1} + R_{i,2} + \cdots + R_{i,T}}{T} = \frac{1}{T}\sum_{t=1}^{T} R_{i,t}$$

Assumes the amount invested at the beginning of each period is the same.

3.3 Geometric Mean Return

$$\bar{R}{G,i} = \sqrt[T]{\prod{t=1}^{T}(1 + R_{i,t})} - 1$$

Example:

$$\sqrt[3]{(1-0.50)(1+0.35)(1+0.27)} - 1 \approx -0.0500 \quad \text{(more precise)}$$

The geometric mean is always less than or equal to the arithmetic mean.
The two are equal only when there is no variability in the observations.

3.4 Harmonic Mean

$$\bar{X}H = \frac{n}{\displaystyle\sum{i=1}^{n}\frac{1}{X_i}}, \quad X_i > 0 \text{ for } i = 1, 2, \ldots, n$$

Example: for ${1, 2, 3, 4, 5, 6, 1000}$, $\bar{X}_H = 2.8560$ (outlier has less impact).

Properties:

  • Measure of central tendency in the presence of outliers
  • Appropriate for averaging ratios (“amount per unit”)
  • Works only for non-negative numbers

Key inequality:

$$\bar{X}_A \geq \bar{X}_G \geq \bar{X}_H$$

$$\bar{X}_A \times \bar{X}_H = \bar{X}_G^2$$

Mean speed example (travel from A to B at speed $v_1$, return at speed $v_2$):

$$\text{mean speed} = \frac{2}{\dfrac{1}{v_1} + \dfrac{1}{v_2}} = \frac{2v_1 v_2}{v_1 + v_2}$$

3.5 Trimmed and Winsorized Means

MethodDescription
Trimmed meanRemoves a small defined percentage of the most extreme observations to limit outlier effects
Winsorized meanReplaces extreme observations with the boundary values to limit outlier effects

Both reduce outlier impact while retaining all data points (Winsorized) or sample size (Trimmed).